Run Independent Component Analysis on gene expression

Run fastica algorithm from the ica package for ICA dimensionality reduction. Wrapper for Seurat's RunICA function, with additional significant gene identification step.

runICA(
  object,
  assay = DefaultAssay(object),
  features = NULL,
  max_cells = 20000,
  verbose = T,
  ...
)

Arguments

object

Seurat object

assay

name of assay to use for ICA analysis. Expression data from 'scale.data' slot will be used.

features

Features to compute ICA on. If not specified, all features present in `object` are used (not recommended, ICA is computationally expensive).

max_cells

Max number of cells to run ICA on. If number of cells in `object` exceeded `max_cells`, object is subsampled prior to running ICA. Default is 20000.

verbose

Print progress. Default is T.

...

Additional parameters passed to Seurat::RunICA(...)

Value

seurat object with significant genes stored in "misc" slot of ICA reduction slot.

References

https://nmikolajewicz.github.io/scMiko/articles/Module_Detection.html

See also

RunICA for Seurat's independent component analysis, getICAGenes for significant ICA gene identification.

Author

Nicholas Mikolajewicz